Combining residence time and isotopic tracers to better understand groundwater reservoir and flows in a karst thermal aquifer
Résumé
Karst hydrosystems are complex systems but often undergo high anthropogenic pressure on their water resources in Mediterranean area since it is shared by many actors. The addition of thermal and/or marine components in these systems makes the interpretation of classical methods more difficult. The thermal karst aquifer of Thau Basin (South of France) illustrates well the complexity of such underpressured karst hydrosystems. In the Balaruc-les-Bains area, groundwater results from the convergence and mixing of (i) young and cold karst water, (ii) old, hot and mineralized thermal water and (ii) marine water (Thau Lagoon and/or seawater). In the framework of the Dem'Eaux Thau CPER/FEDER project (2017-2022), we propose to combine tracers of water-rock interaction processes (Sr, Li) and residence time tracers (<sup>4</sup>He, <sup>14</sup>C, <sup>36</sup>Cl) to better understand the origin and mean residence times of flow that takes place in the system, with a focus on the thermal water. In particular, the originality of this work was to constraint geological and hydrogeological informations using natural tracers and to calibrate He dating using <sup>14</sup>C ages.</p><p>The combination of Si geothermometer and isotopic Sr signature (<sup>87</sup>Sr/<sup>86</sup>Sr) indicates that the thermal reservoir is located in the Jurrassic carbonate formation until 2000m depth. In addition Li concentrations show the existence of deep flows from granitic bed-rock. These new geochemical results allowed to better constraint the location of the thermal reservoir on the geological 3D map of the area and points out the major role of a local fault. However, there is significant uncertainty on the porosities of this reservoir impacting He age dating method. We used Carbon-14 dating in a deep karst well to constrain <sup>4</sup>He ages and therefore determine the reservoir porosity. In parallel, the identification and quantification of the thermal flux by Li concentrations, allowed us to correct the <sup>4</sup>He concentrations, and to propose a residence time of the thermal water of several thousands of years (10 000 to 50 000 years) which were consistent with the <sup>36</sup>Cl results. Thermal water subsequently feed a shallow reservoir (100 – 300 m) through local fractures and mix with variable proportions of recent karst flows.
